Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference

Matthias Seeger, Hannes Nickisch
Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, PMLR 15:652-660, 2011.

Abstract

We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper & Winther, 2005) with covariance decoupling techniques (Wipf & Nagarajan, 2008; Nickisch & Seeger, 2009), it runs at least an order of magnitude faster than the most common EP solver.

Cite this Paper

BibTeX
@InProceedings{pmlr-v15-seeger11a, title = {Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference}, author = {Seeger, Matthias and Nickisch, Hannes}, booktitle = {Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics}, pages = {652--660}, year = {2011}, editor = {Gordon, Geoffrey and Dunson, David and Dudík, Miroslav}, volume = {15}, series = {Proceedings of Machine Learning Research}, address = {Fort Lauderdale, FL, USA}, month = {11--13 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v15/seeger11a/seeger11a.pdf}, url = {https://proceedings.mlr.press/v15/seeger11a.html}, abstract = {We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper & Winther, 2005) with covariance decoupling techniques (Wipf & Nagarajan, 2008; Nickisch & Seeger, 2009), it runs at least an order of magnitude faster than the most common EP solver.} }
Endnote
%0 Conference Paper %T Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference %A Matthias Seeger %A Hannes Nickisch %B Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2011 %E Geoffrey Gordon %E David Dunson %E Miroslav Dudík %F pmlr-v15-seeger11a %I PMLR %P 652--660 %U https://proceedings.mlr.press/v15/seeger11a.html %V 15 %X We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper & Winther, 2005) with covariance decoupling techniques (Wipf & Nagarajan, 2008; Nickisch & Seeger, 2009), it runs at least an order of magnitude faster than the most common EP solver.
RIS
TY - CPAPER TI - Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference AU - Matthias Seeger AU - Hannes Nickisch BT - Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics DA - 2011/06/14 ED - Geoffrey Gordon ED - David Dunson ED - Miroslav Dudík ID - pmlr-v15-seeger11a PB - PMLR DP - Proceedings of Machine Learning Research VL - 15 SP - 652 EP - 660 L1 - http://proceedings.mlr.press/v15/seeger11a/seeger11a.pdf UR - https://proceedings.mlr.press/v15/seeger11a.html AB - We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper & Winther, 2005) with covariance decoupling techniques (Wipf & Nagarajan, 2008; Nickisch & Seeger, 2009), it runs at least an order of magnitude faster than the most common EP solver. ER -
APA
Seeger, M. & Nickisch, H.. (2011). Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 15:652-660 Available from https://proceedings.mlr.press/v15/seeger11a.html.

Related Material